Deciphering the Molecular Mechanisms Underlying Cochlea Development in Mice using Next- Generation Sequencing Technologies and Comprehensive Computational Approaches

Abstract

Hearing is mediated by the specialized sensory epithelium known as the organ of Corti, which is located in the cochlea of the inner ear. One hallmark of the cochlea is the ability to discriminate between different sound frequencies. This characteristic is based on the tonotopic organization of the organ of Corti, where different frequencies are detected by the sensory cells depending on their position along the longitudinal axis.

In this thesis, we first applied single-cell RNA-sequencing (scRNA-seq) from two distinct time points to investigate the molecular mechanisms of tonotopic patterning during embryonic development using a computational framework. We proposed two biological hypotheses regarding the tonotopic patterning during cochlear duct extension and tested the hypotheses by leveraging the scRNA-seq datasets. Our findings suggested that spatial identity in the developing cochlea was conferred by morphogens rather than a cell division-associated mechanism. Subsequently, the 3D anatomical structure of the developing cochlea was reconstructed from scRNA-seq data to identify morphogens mediating longitudinal patterning. Opposing gradients of the retinoic acid (RA) and sonic hedgehog (SHH) were found along the tonotopic axis during development. Functional interrogation using mouse cochlear explants supported the notion that both RA and SHH jointly function to specify the tonotopic axis.

Next, we provided a comprehensive computational pipeline to identify the regulatory landscape controlling the differentiation of the organ of Corti. By utilizing single-cell assay for transposase accessible chromatin using sequencing (scATAC-seq) and scRNA-seq techniques from genetically labeled mouse hair cells and supporting cells after birth, we predicted cell type-specific functions of developmental transcription factors and reconstruct gene regulatory networks. Comparative analysis determined 20 hair cell-specific activators and repressors, along with their downstream target genes. Clustering of target genes revealed related transcription factors and inferred their developmental functions. Furthermore, spatial reconstruction of transcriptional and chromatin accessibility trajectories suggested that the formation of the cell type specific chromatin accessibility landscape is lagging behind their transcriptional identity.

Lastly, we employed our computational pipeline in conjunction with laboratory experiments to identify RA regulated genes in organ of Corti development. RA signaling in the cochlea is mediated via binding of the ligand to the RARA receptor that functions as transcription factor. Using computational approaches, we explored the role of RARA in cochlear hair cell differentiation. Among the genes that are regulated by RARA, we found Lfng, a known supporting cell marker. Analyzing scRNA-seq and scATAC-seq data, we found that in absence of RA the function of RARA is to repress the expression of supporting cell specific genes in postnatal hair cells. Interestingly, ectopic RA was found to induce the expression of Lfng, suggesting that binding of RA renders RARA into a transcriptional activator, which likely is the case during embryonic development of the cochlea.